Merge pull request #845 from IanCa/develop

Remove some URL calls in test code, try to update token for workflow

.github/workflows/ci.yaml

@@ -64,11 +64,15 @@ jobs:
           # exit-zero treats all errors as warnings. The GitHub editor is 127 chars wide
 
       - name: Test with unittest
+        env:
+          HED_GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         run: |
-          HED_GITHUB_TOKEN=${{ secrets.HED_GITHUB_TOKEN }} coverage run -m unittest
+          coverage run -m unittest
 
       - name: Run spec_test coverage
-        run: HED_GITHUB_TOKEN=${{ secrets.HED_GITHUB_TOKEN }} coverage run --append -m unittest spec_tests/test_errors.py
+        env:
+          HED_GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: coverage run --append -m unittest spec_tests/test_errors.py
         continue-on-error: true
 
       - name: Archive code coverage results

hed/models/base_input.py

@@ -135,6 +135,7 @@ def _indexed_dict_from_onsets(onsets):
 
         return indexed_dict
 
+    # This would need to store the index list -> So it can optionally apply to other columns on request
     @staticmethod
     def _filter_by_index_list(original_series, indexed_dict):
         new_series = pd.Series(["n/a"] * len(original_series), dtype=str)
@@ -264,7 +265,7 @@ def to_csv(self, file=None):
             - Cannot open the indicated file
         """
         dataframe = self._dataframe
-        csv_string_if_filename_none = dataframe.to_csv(file, '\t', index=False, header=self._has_column_names)
+        csv_string_if_filename_none = dataframe.to_csv(file, sep='\t', index=False, header=self._has_column_names)
         return csv_string_if_filename_none
 
     @property

hed/tools/analysis/key_map.py

@@ -190,7 +190,9 @@ def _update(self, base_df):
             next_pos += pos_update
         if row_list:
             df = pd.DataFrame(row_list)
-            self.col_map = pd.concat([self.col_map, df], axis=0, ignore_index=True)
+            # Ignore empty col_map to suppress warning
+            col_map = self.col_map if not self.col_map.empty else None
+            self.col_map = pd.concat([col_map, df], axis=0, ignore_index=True)
 
     def _handle_update(self, row, row_list, next_pos):
         """ Update the dictionary and counts of the number of times this combination of key columns appears.

hed/tools/bids/bids_dataset.py

@@ -8,9 +8,6 @@
 from hed.tools.bids.bids_file_group import BidsFileGroup
 
 
-LIBRARY_URL_BASE = "https://raw.githubusercontent.com/hed-standard/hed-schemas/main/library_schemas/"
-
-
 class BidsDataset:
     """ A BIDS dataset representation primarily focused on HED evaluation.
 

hed/tools/remodeling/dispatcher.py

@@ -11,6 +11,12 @@
 from hed.tools.remodeling.operations.valid_operations import valid_operations
 from hed.tools.util.io_util import clean_filename, extract_suffix_path, get_timestamp
 
+# This isn't supported in all versions of pandas
+try:
+    pd.set_option('future.no_silent_downcasting', True)
+except pd.errors.OptionError:
+    pass
+
 
 class Dispatcher:
     """ Controller for applying operations to tabular files and saving the results. """
@@ -193,7 +199,10 @@ def prep_data(df):
             df (DataFrame) - The DataFrame to be processed.
 
         """
-        return df.replace('n/a', np.NaN)
+        result = df.replace('n/a', np.NaN)
+        # Comment in the next line if this behavior was actually needed, but I don't think it is.
+        # result = result.infer_objects(copy=False)
+        return result
 
     @staticmethod
     def post_proc_data(df):

tests/tools/bids/test_bids_dataset.py

@@ -87,15 +87,7 @@ def test_validator_types(self):
         self.assertFalse(issues, "BidsDataset with participants and events validates")
 
     def test_with_schema_group(self):
-        base_version = '8.2.0'
-        library1_url = "https://raw.githubusercontent.com/hed-standard/hed-schemas/main/" + \
-                       "library_schemas/score/hedxml/HED_score_1.0.0.xml"
-        library2_url = "https://raw.githubusercontent.com/hed-standard/hed-schemas/main/" + \
-                       "library_schemas/testlib/hedxml/HED_testlib_1.0.2.xml"
-        schema_list = [load_schema_version(xml_version=base_version),
-                       load_schema(library1_url, schema_namespace="sc"),
-                       load_schema(library2_url, schema_namespace="test")]
-        x = HedSchemaGroup(schema_list)
+        x = load_schema_version(["8.2.0", "sc:score_1.0.0", "test:testlib_1.0.2"])
         bids = BidsDataset(self.library_path, schema=x, tabular_types=["participants"] )
         self.assertIsInstance(bids, BidsDataset, 
                               "BidsDataset with libraries should create a valid object from valid dataset")

tests/tools/bids/test_bids_file_group.py

@@ -1,6 +1,6 @@
 import os
 import unittest
-from hed.schema.hed_schema_io import load_schema
+from hed.schema.hed_schema_io import load_schema_version
 from hed.tools.analysis.tabular_summary import TabularSummary
 from hed.tools.bids.bids_file_group import BidsFileGroup
 
@@ -31,8 +31,7 @@ def test_constructor(self):
 
     def test_validator(self):
         events = BidsFileGroup(self.root_path)
-        hed = 'https://raw.githubusercontent.com/hed-standard/hed-schemas/main/standard_schema/hedxml/HED8.0.0.xml'
-        hed_schema = load_schema(hed)
+        hed_schema = load_schema_version("8.0.0")
         validation_issues = events.validate_datafiles(hed_schema, check_for_warnings=False)
         self.assertFalse(validation_issues, "BidsFileGroup should have no validation errors")
         validation_issues = events.validate_datafiles(hed_schema, check_for_warnings=True)

tests/tools/util/test_data_util.py

@@ -69,7 +69,7 @@ def test_get_new_dataframe(self):
         self.assertIsInstance(df_new1, DataFrame)
         self.assertEqual(len(df_new1), 87, "get_new_dataframe should return correct number of rows")
         self.assertEqual(len(df_new1.columns), 4, "get_new_dataframe should return correct number of rows")
-        df_new.iloc[0]['type'] = 'Pear'
+        df_new.loc[0, 'type'] = 'Pear'
         self.assertNotEqual(df_new.iloc[0]['type'], df_new1.iloc[0]['type'],
                             "get_new_dataframe returns a new dataframe")